As applications of mineral magnetism to the study of the Earth and other planetary bodies diversify, comprehensive mineral phase identification of multi-component assemblages is increasingly important. The accuracy with which the stoichiometry of the magnetic mineral phases, their grain size distributions and concentrations are determined sets the limits of mineral magnetism interpretations.

Despite the ubiquitous occurrence of iron-bearing ferromagnetic (s.l.) minerals on earth and extraterrestrial bodies (e.g., Lagroix et al., 2016) they are almost exclusively present as accessory minerals in rocks and sediments, with the exception, for example, of iron-ore deposits and banded-iron formations. Ferromagnetic (s.l.) minerals often go undetected by traditional mineral phase speciation analyses such as XRD or Mössbauer spectroscopy because concentrations in bulk material are typically less than a few weight percent. Mineral magnetism, on the other hand, can detect magnetic minerals at concentrations as low as 1 ppm. It has been shown that physical mineral separation, e.g., magnetic extracts or heavy liquid separation, does not comprehensively isolate the entire ferromagnetic (s.l.) mineral assemblage (e.g., Lagroix et al., 2004; Wang et al., 2013) and chemical separation techniques, such as CBD or oxalate, enable to quantify dissolvable Fe concentrations but do not identify the iron-bearing mineral phases (e.g., Mehra and Jackson, 1960; Vidic et al., 2000; van Oorschot and Dekkers, 2001; van Oorschot et al., 2002). Therefore, mineral magnetic analyses on bulk material provide the greatest potential for a comprehensive characterization of iron-bearing ferromagnetic (s.l.) minerals.

Remanence acquisition or demagnetization curves and field-dependence of induced magnetization probed through hysteresis loops and first-order reversal curve (FORC) distributions are the dominant magnetic data currently being numerically unmixed to characterize the magnetic mineral components (Robertson and France, 1994; Carter-Stiglitz et al., 2001; Egli, 2003, 2013; Fabian, 2003; Heslop and Dillon, 2007; Lascu et al., 2010, 2015; Heslop, 2015; Church et al., 2016; Fabian et al., 2016; Maxbauer et al., 2016b; Zhang et al., 2016). While important and useful results have emerged, the currently applied methods present important limitations. The commonly acquired magnetic data listed above are generally acquired at room temperature. Consequently, all iron-bearing mineral phases ordering magnetically below room temperature (e.g., siderite, titanium-rich (>80% Ti) compositions of both magnetite-ulvöspinel and hematite-ilmenite series, lepidocrocite, chromite, daubrelite) will go undetected. Furthermore, for remanence data, fine particles smaller than the superparamagnetic to stable single domain grain size boundary at room temperature will also go undetected. Induced magnetization data is inclusive of all minerals but differences over three orders of magnitude in saturation magnetizations heavily biases hysteresis loop derived data toward strongly ferrimagnetic minerals (e.g., magnetite, see for instance Frank and Nowaczyk, 2008). Decomposing induced magnetization data requires a priori knowledge of basis functions of the individual components (Carter-Stiglitz et al., 2001). These limitations compromise significantly the comprehensiveness of magnetic mineral assemblage characterization and consequently mineral magnetism-based interpretations of geological and biogeochemical processes occurring on Earth and other planetary bodies.

Here, we present a new experimental method using existing equipment for decomposing the mineral magnetic behavior of multi-phase assemblages. The method is validated on synthetic and natural samples. Finally, new frontiers for earth and planetary sciences through mineral magnetism are discussed.

The new experimental protocol for separating the low-temperature magnetic behavior of natural geological samples utilizes a MPMS XL-5 EverCool, a commercially available cryogenic magnetometer from Quantum Designs ©. The configuration of the system used includes the ultra-low field, AC and EDC (External Device Control) options, which are needed to run the experimental protocol. The three options enable to measure the residual field of the superconducting magnet once it has been charged and set back to zero field, and to compensate for the residual field by applying a directionally opposing yet equal field. By compensating for the residual field of the superconducting magnet, a zero field inferior or equal to 500 nT is achievable using the Quantum Design MPMS-specific fluxgate. The protocol is a remanent magnetization-based experiment, which can be applied to demagnetize an acquired isothermal or thermal remanent magnetization (IRM or TRM) at any temperature over the instrument 2–400 K temperature range. It is analogous to the protocol suggested in Guyodo et al. (2006) and shares some common features with those of Carter-Stiglitz et al. (2006) and Lascu and Feinberg (2011). The distinguishing feature here is that the isothermal demagnetization step is achieved with the MPMS itself instead of an external alternating field demagnetizer. Advantageously, this eliminates the need for multiple instruments and excessive sample handling and more importantly enables the complete or partial demagnetization of IRMs or TRMs acquired at temperatures other than room temperature. An isothermal demagnetization results from charging the superconducting magnet from one set field to a lower set field using the system's oscillation superconducting magnet charging mode, which produces a series of DC fields of changing polarity and amplitude with a set frequency, resulting in a demagnetization. The resulting DC demagnetization is very similar to that of an AF demagnetizer. To date, the no-overshoot mode was routinely used to charge the superconducting magnet back to zero field in order to preserve the acquired IRM or TRM. The aim of Figure 1 is to illustrate the differences between the two modes by tracing the time evolution of the voltage read by the internal MPMS voltmeter. The voltmeter readings are not direct readings of the current trapped in the superconducting magnet nor its polarity. In the no-overshoot mode, the DC current supplied to the magnet varies rapidly until it reaches 70% of the difference between the initial and target value, then varies slowly to reach the final value without overshooting the target. In the oscillation mode, the current provided to the magnet oscillates in decreasing amplitude to reach the final value. The new experimental protocol was first publicly presented at the EGU General Assembly in 2014 (Lagroix et al., 2014). The use of the MPMS's oscillation mode for demagnetization was mentioned thereafter in the appendix of Maxbauer et al. (2016a).

The MPMS isothermal demagnetization protocol was tested and validated on three mono-mineral synthetic powders and one sample composed of a mixture of the three components. Component 1 is a synthetic magnetite powder [Wright Industries Inc. (W3006)] commercialized as multi-domain ~2–3 μm particles. W3006 magnetite has been used in several previous studies (Dunlop, 2002; Kosterov, 2002; Liu et al., 2002; Smirnov and Tarduno, 2002; Yu et al., 2002; Till et al., 2010). Component 2 is a hematite powder with approximate particle size of 10–40 nm based on TEM images synthesized from ferrihydrite following Schwertmann and Cornell (1991) by David Burleson in the laboratory of Prof. R. Lee Penn in the Chemistry Department at the University of Minnesota. Component 3 is a goethite powder composed of micrometer size needle crystals (30 × 350 nm) synthesized following the method of Schwertmann and Cornell (1991) and previously used in Guyodo et al. (2003). Staking the three components with 4.2 mg of magnetite sandwiched in between 74.8 mg of goethite and 102 mg of hematite formed the “mixture” sample. Magnetite, goethite and hematite components represent, respectively, 2.3, 41.3, and 56.3 wt.% of the mixed sample. Further demonstrations are provided on a natural loess sample of the penultimate glacial period (Saalian glacial or marine oxygen isotope stage 6) from the Dolni Vestonice sequence in the Czech Republic (Antoine et al., 2013; Fuchs et al., 2013) and a commercially available hematite sample from The Nature Company (www.naturesowndesigns.com), which is used as a remarkably stable MD magnetite + hematite standard. All samples were prepared in gelatin capsules and when necessary completed with cotton to fill the volume of the capsules. For measurements, the gelatin capsules were inserted in clear plastic straws, following common practices.

The magnetic behavior of the individual components and of the mixture sample is characterized at room temperature (RT) and at low-temperature (LT) (Figure 2, Table 1). Room temperature hysteresis loops were measured on a vibrating sample magnetometer (VSM) in a 1.5 T maximum field. The behavior of a RT saturation isothermal remanent magnetization (SIRM) acquired in a 2.5 T field was monitored while cycling from 300 to 10 K and back to 300 K at a sweep rate of 5 K/min, which is the same rate used throughout. This protocol is commonly referred to as a RT-SIRM experiment even though saturation of the IRM is not necessarily achieved. The behavior of a LT-SIRM acquired in a 2.5 T field at 10 K after cooling the sample from 300 to 10 K in zero field (ZFC) or in a 2.5 T field (FC) is monitored upon warming from 10 to 300 K. This protocol is routinely referred to as a ZFC-FC remanence experiment.

In order to test the demagnetization efficiency of the superconducting magnet's oscillation mode, the individual components and mixture samples were subjected to a step-wise demagnetization of an IRM acquired in a 2.5 T field, followed by a step-wise IRM acquisition up to 2.5 T and a one-step demagnetization by charging the superconducting magnet down in the oscillation mode from 300 to 0 mT (Figure 3). The experimental sequence programmed through the MultiVu software is provided as Supplementary Material.

The experimental protocol followed to generate the data of Figure 4, aiming at decomposing the magnetic behavior of the individual components within the mixture sample following the acquisition of an enhanced RT-SIRM, is provided as Supplementary Material. The sequence is as follows:

Acquisition of an IRM in a 2.5 T field at 300 K (RT-SIRM).

The same experimental sequence is run on a natural sample in Figure 5. The RT-SIRM is labeled as enhanced because the sample is cooled from 400 to 300 K in a strong applied field. If any goethite is present, this RT-SIRM pre-treatment effectively imparts a TRM on the goethite population as it cools through its Néel temperature (393 K for stoichiometric goethite) in the presence of a magnetic field. Finally, in Figure 6, we illustrate the new capability by demagnetizing an IRM acquired at 20 K in the case of a ZFC-FC experiment. The experimental sequence executed to generate the data of Figure 6 is provided as Supplementary Material.

One timeless challenge in rock magnetic studies, inclusive of paleomagnetism and environmental magnetism, is decomposing a sample's bulk magnetic behavior into its individual magnetic mineral components. For the study of the Earth's magnetic field and its derived applications for paleogeographic, tectonic and geodynamic reconstructions, the breadth of useful mineral stoichiometry and grain size is restricted to those ordering magnetically at a temperature well above ambient temperature and having a grain size stable magnetically over geological time scales. In contrast, for environmental magnetism studies the breadth of useful mineral stoichiometry and grain size is unrestricted. The accuracy of magnetism-based interpretations of climate, environmental and biogeochemical processes and change is dependent, firstly, on a comprehensive identification of all magnetic mineral compositions, no matter their magnetic ordering temperature or particle size. The other two important determinations are the concentration and grain size distribution for each mineral composition.

The magnetic data acquired on the three components and the mixture samples used in this study (Figure 2, Table 1) illustrate fairly well the difficulties in obtaining comprehensive chemical composition, and (semi-)quantitative grain size and concentration estimates even for the most common of iron oxides (magnetite and hematite) and oxyhydroxides (goethite). Scalar quantities derived from hysteresis parameters and their ratios are the most accessible means of characterizing relative variations in mineral composition (e.g., S-ratio, HIRM), mean magnetic grain size (e.g., M_RS/M_S, χ/M_S), and concentration (M_S, M_RS, χ, χ_f, χ_HF) (see review by Liu et al., 2012). The mixture sample, despite being constituted of 97.7 wt% hematite + goethite and only 2.3 wt% magnetite has hysteresis parameters which fall on the magnetite SD-MD mixing lines of Dunlop (2002). Within a natural bulk sample containing paramagnetic and diamagnetic minerals, and in a maximum field of 1.5 T, the magnetization of component 3 (goethite) as well as 75% of the magnetization of component 2 (nano-hematite) would be lost in the high field slope corrected magnetization. Scalar parameters acquired in low to moderately high fields will thus provide an incomplete picture of the ensemble of environmentally significant iron bearing minerals, heavily biased toward magnetite. While this statement is common knowledge, it is often disregarded in reports and reviews of environmental magnetism studies.

Unmixing and principal component analysis of spectrum-type data such as remanence curves, hysteresis loops and FORC distributions can more readily provide evidence of multiple components. These approaches are increasingly undertaken to characterize changes in contributions to the total remanent magnetization and in magnetic grain size of sedimentary archives of various origins from past and present continents and oceans (e.g., Abrajevitch and Kodama, 2011; Lascu et al., 2012; Chen et al., 2014; Nie et al., 2014; Hyland et al., 2015; Fabian et al., 2016; Maxbauer et al., 2016c; Zhang et al., 2016). However, obtaining mineral compositions of an ensemble of components via these methods is only possible indirectly through a series of assumptions or a priori knowledge. In addition, the fact that these analyses are isothermal, (i.e., investigating changes in magnetization as a function of field strength at a single temperature) and that the analysis temperature is almost always room temperature, there is a high probability of incomprehensiveness of the characterization results, similarly to scalar derived hysteresis parameters discussed above. The thermal fluctuation tomography method proposed by Jackson et al. (2006) addresses the limitation of exclusion of particles with SP behavior from room temperature-derived coercivity distribution, and provides a means of characterizing grain size distributions from variable low-temperature remanence coercivity spectrums. Unfortunately, the method has been seldom applied (e.g., Nie et al., 2013; Wang et al., 2013) likely due to the time consuming and relatively costly nature of the analysis.

The most direct means of identifying the chemical composition of magnetic minerals is through thermomagnetic data and the observation of magnetic ordering transitions (Curie and Néel temperature) or other characteristic crystallographic or magnetic transitions (e.g., Verwey, Morin and Besnus transitions, spin glass transition). Many transitions can be observed below 400 K. In Figure 2, for instance, we observe the drastically different low temperature behavior of the three components for both the RT-SIRM and ZFC-FC experiments. This said, the pitfalls due to the overwhelming magnetization of component 1 in the mixture, observed and discussed above, remain even for low-temperature thermomagnetic data. Low-temperature cycling of a RT-SIRM fails to detect any change in intensity that might hint to an additional component other than magnetite (Figure 2K). The same is also true for the ZFC data. The only hint of a second component comes from the small inflection around 220 K in the FC warming curve (Figure 2L). Because heating induced alterations are absent over these temperature ranges, multiple experiments can be conducted on the same sample. With the added new possibility of partially demagnetizing (at any temperature between 2 and 400 K) remanences acquired in fields up to 5, 7, or 14 T and at temperatures as low as 2 K using the same equipment, experimental decomposition of a bulk sample's thermomagnetic behavior, as demonstrated and validated herein (Figures 4–6), becomes possible. The new experimental method presented here is currently the only means of obtaining a comprehensive identification of all mineral species of an ensemble of magnetic mineral within a bulk sample, with a minimum of assumptions and no a priori knowledge.

FL and YG both contributed to the experimental development, the data analysis and the writing of the manuscript.